基于灭点的径向畸变自动校正方法

镜头畸变影响着三维重建、几何量测等工作的质量。本文根据图像中的灭点几何约束条件，提出一种基于灭点的弱径向畸变自动校正方法。为避免包含畸变中心参数的非线性模型优化结果的不稳定性，本文在求解径向畸变系数之后进一步对畸变中心进行优化。首先根据灭点几何约束条件建立关于灭点与径向畸变系数的非线性模型，使用LM算法估计灭点坐标与径向畸变参数，然后根据质量评价准则对畸变中心和径向畸变系数进一步迭代优化；最后，通过真实图像对该方法的可行性进行分析验证。本文充分利用图像中的灭点属性，提出一种新的镜头径向畸变校正方法。实验结果表明，该方法能够有效的对径向畸变进行校正，并克服了传统校正方法的不稳定性。

An Automatic Approach for Radial Distortion Correction from Vanishing Points

Lens distortion plays an essential role which affects quality results of several photogrammetric tasks such as: three-dimension reconstruction and geometric measurement. According to the geometric constraint of vanishing points in a single image, we present an automatic correction of weak radial distortion using vanishing points. To avoid the instability of the results caused by the nonlinear model included in the distortion center, we further optimize the distortion center after estimating the radial distortion. First, according to the geometric constraint of vanishing points, establish a nonlinear model for parameters of vanishing points and radial distortion, which will be estimated by LM algorithms, then further optimize the distortion center and radial distortion by a suitable quality measurement. Finally, the real experiments are presented and discussed. We take full advantage of geometric constraint of vanishing points. We present a new approach for radial distortion correction. The results of real experiments indicate that the approach can effectively correct radial lens distortion, while effectively overcame the instability of the traditional methods.